6 BULLETIN 62, U. S. DEPABTMEXT OF AGBICULTUBE. 
cotton-growing and cott on-manufacturing regions of the United 
States, and these experts without exception declared the bales to he 
excellently selected. 
While this cotton is now well through the mill, so that the success 
of the experiment is assured, it is not yet possible to draw more than 
tentative and approximate conclusions on a number of points. It 
seems safe, however, to make the following statements. 
PERCENTAGE OF WASTE. 
The two classes of cotton, Western Upland and Atlantic States 
Upland, have yielded a visible waste of slightly different weight and 
character, the average difference in the percentage of waste being 
between 1 and 2 per cent, taking all the grades into consideration. 
This difference obtained in the mill has been paralleled by carefully 
made hand separations. In the hand separations, the average 
difference in waste was about If per cent. On the whole, as would 
be expected, the differences are considerably greater in the lower 
grades than in the higher grades. The highest difference so far noted 
was the following, but so large a difference appears altogether excep- 
tional. 
Visible waste {hand separated). — Atlantic States Upland Good 
Ordinary, 12.49 per cent; Western Good Ordinary, 7.80 per cent. 
Mill waste {visible). — The corresponding minimum difference was as 
follows: Atlantic States Upland Good Ordinary, 12.57 percent; West- 
ern Good Ordinary, 10.08 per cent. 
VALUATION OF THE WASTE. 
The value of the visible waste from the various grades has yet to be 
determined, but from its character there can be little doubt that the 
valuation figures for the waste of the two classes of cotton will be 
approximately equal, weight for weight. 
TENSILE STRENGTH OF THE YARN. 
Preliniinar} 7 and approximate figures have been obtained con- 
cerning the tensile strength of the yarns. These tests show the yarn 
from the two classes of cotton to be about equal in strength. 
As regards the relative amount of visible waste in the different 
grades, the figures are found to be more consistent than might have 
been expected. The mill waste in the experiments to date varies 
from about 4 per cent in Middling Fair to about 1 1 per cent in Good 
Ordinary, and the various official grades tested fall into line with 
something approaching mathematical uniformity, as will be seen by 
examination of the following graph (fig. 1). 
